Thermoresponsive Cellulosic Hydrogels with Cell-Releasing BehaviorSiew P. Hoo,†,‡ Fatemeh Sarvi,§ Wai Ho Li,∥ Peggy P.Y. Chan,*,‡,⊥ and Zhilian Yue*,‡,#

†Department of Chemical Engineering, Monash University, Australia‡MicroNanophysics Research Laboratory, School of Applied Science, RMIT University, Australia§Department of Mechanical and Aerospace Engineering, Monash University, Australia∥Monash Vision group, Monash University, Australia⊥Melbourne Centre for Nanofabrication, Clayton, Australia#ARC Centre of Excellence for Electromaterials Science, Intelligent Polymer Research Institute, University of Wollongong, Australia

ABSTRACT: Here we report the preparation and character-ization of thermoresponsive cellulosic hydrogels with cell-releasing behavior. Hydroxypropyl cellulose (HPC) wasmodified with methacrylic anhydride (MA). The resultantmacromonomer, HPC-MA, retains the characteristic thermor-esponsive phase behavior of HPC, with an onset temperatureof 36 °C and a lower critical solution temperature (LCST) of37−38 °C, as determined by turbidity measurement.Homogenous HPC-MA hydrogels were prepared by UV-cross-linking the aqueous solutions of the macromonomer atroom temperature, and characterized by water contact angleand swelling ratio measurements, and dynamic mechanical analysis. These hydrogels exhibit temperature-dependent surfacehydrophilicity and hydrophobicity, equilibrium water content as well as mechanical properties. Cell-releasing characteristics weredemonstrated using African green monkey kidney cell line (COS-7 cells) and murine-derived embryonic stem cell line (Oct4b2).By reducing temperature to 4 °C, the cultivated cells spontaneously detached from the hydrogels without the need of trypsintreatment. These unique properties make our HPC-MA hydrogels potential substrates for cell sheet engineering.

KEYWORDS: thermoresponsive hydrogels, cellulose, lower critical solution temperature, murine-derived embryonic stem cells,cell sheet engineering

■ INTRODUCTIONThere has been extensive research into intelligent substratesthat allow for spontaneous cell harvesting in response to anenvironmental stimulus.1−3 A typical example is the thermor-esponsive culture dishes prepared by grafting nanometer-sizedpoly(N-isopropylacrylamide) (PNIPAAm), or its copolymers,onto commercial tissue culture polystyrene dishes.4 PNIPAAmundergoes a distinct phase transition from a dehydrated,collapsed structure to a highly hydrated, extended structureacross a lower critical solution temperature (LCST) of 32 °C.5,6

Such thermosensitive nature allows controlled alteration ofsurface hydrophobicity and hydrophilicity of the PNIPAAm-grafted dishes and, consequently, cell adhesion and detachment,by a small change in temperature. When temperature is belowthe LCST, cells cultivated on the substrate spontaneouslydetach as intact cell sheets with retained cell-to-cell junctionsand extracellular matrix proteins, without the use of traditionalmethods involving enzymatic digestion or mechanical treat-ment.7,8 Cell sheet engineering represents a novel scaffold-freeapproach suited for the regeneration of cell-dense tissues suchas skin,9 corneal,10 cardiac,11,12 esophageal, and hepatic13

tissues.Efforts have been made into extending the thermoresponsive

culture surfaces with thermoresponsive hydrogels for cell sheet

engineering. The highly hydrated structures and soft-tissue-likebiomechanical properties of hydrogels and, more importantly,their capability for sustained delivery of bioactive molecules canprovide a more physiologically relevant environment for thecultivated cells. For example, interpenetrating nanocompositehydrogels have been prepared from hectorite clay nanoparticles,PNIPAAm and alginate. The presence of alginate has shown tosignificantly accelerate cell detachment, as a result of improvedwater penetration in the resulting hydrogels.14 AlthoughNIPAAm have been demonstrated to be a promising substratefor cell sheet engineering, NIPAAm is not derived fromrenewable resources, and NIPAAm does induce cellularcytotoxicity at physiological temperature.15

A number of studies have attempted to developed newthermoresponsive substrates to engineer cell sheets. Forexample, hydroxybutyl chitosan,16 poly(N-isopropylacryla-mide)/clay nanocomposite hydrogel,17 methylcellulose/colla-gen hydrogel,18 poly(N-isopropylacrylamide-co-acrylic acid)-b-poly(L-lactic acid),19 and elastic protein-based polymer.20 Somestudies have attempted to engineer cell sheets using non-

Received: March 13, 2013Accepted: June 4, 2013Published: June 4, 2013

Research Article

www.acsami.org

© 2013 American Chemical Society 5592 dx.doi.org/10.1021/am4009133 | ACS Appl. Mater. Interfaces 2013, 5, 5592−5600

thermoresponsive substrate; for example, Guillaume-Gentil etal.21 developed a RGD-modified poly(L-lysine)-graf t-poly-(ethylene glycol) substrate that can release cell sheet uponelectrochemical polarization; Edahiro et al.22 developed aphotoresponsive NIPAAm-nitrospiropyran substrate that re-leases cells after UV irradiation and low temperature washing;Nagai et al.23 developed a salmon atelocollagen fibrillar gel thatreleases cell sheet when subjected to collagenase digestion.Table 1 listed the substrates that have been studied for cellsheet engineering.

Due to the many promising applications of thermoresponsivepolymer, new thermoresponsive polymer systems with newbiochemical properties, new physiochemical properties, lowercytotoxicities, and preferably derived from renewable resources,are still needed to provide multifunctional platforms. Herein,we report the preparation of a thermoresponsive hydrogelbased on hydroxypropyl cellulose (HPC). HPC is a commercialderivative of cellulose, the most abundant renewable poly-saccharide resource; it shows no cytotoxicity and has beenapproved by the Food and Drug Administration (FDA) as anagent for drug delivery applications. Also, HPC is soluble inwater and many organic solvents. This has partly contributed toits use in biomedical and pharmaceutical applications. Moreimportantly, for this study, HPC exhibits a phase transitionfrom isotropic aqueous solution to metastable biphasic systemabove its LCST.24−28 In this work, HPC was modified withbifunctional methacrylic anhydride (MA). The resultingpolymer contains photocross-linkable methacrylate pendantgroup, namely HPC-MA. The thermoresponsive propertiessuch as turbidity, water contact angle, swelling ratio, anddynamic mechanical analysis of these HPC-MA hydrogel wereevaluated. Temperature-modulated cell-releasing characteristicswere studied using COS-7 (African green monkey kidney) cellline and Oct4b2 (murine-derived embryonic stem cell) cell line.

■ MATERIALS AND METHODSSynthesis. All chemicals used were purchased from Sigma-Aldrich,

Australia, unless otherwise stated. Hydroxypropyl cellulose (HPC, Mn= 10 000 g/mol, degree of etherification ∼ 3.4, as determinedpreviously by 1H NMR24) was dehydrated by azeotropic distillation intoluene. The HPC hydrogel precursor was prepared by modifyingHPC with methacrylic anhydride (MA). Briefly, the dehydrated HPC(4.0 g) was dissolved in chloroform (150 mL). MA (4.17 mmol)followed by N,N-dicyclohexylcardodiimide (DCC, 4.17 mmol) wasadded dropwise in chloroform, respectively. The solution was thenstirred for 48 h in the presence of 4-dimethylaminopyrinde (DMAP,0.4 g) as catalyst. The solution was concentrated and precipitated into

diethyl ether. The polymer was then collected, redissolved inchloroform and precipitated into diethyl ether. Finally, the productwas dissolved in water, filtered to remove any insoluble impurities, anddialyzed against deionized water for 72 h before lyophilization in afreeze-dryer (HETO PowerDry PL6000, Thermo Scientific, Australia).It was denoted as HPC-MA and characterized by 1H NMRspectroscopy in CDCl3.

1H NMR (CDCl3, δppm): 0.5−1.5(−CH3−CH−), 1.8−2.0 (CH3−CCH2), 5.5−6.2 (−CCH2),2.5−5.3 (other protons). The degree of modification of HPC-MA isdefined as the number of MA groups per repeating unit, anddetermined by 1H NMR spectroscopy.

Turbidimetry Measurement of HPC-MA in Deionized Water.The turbidity of HPC-MA was investigated by dissolving the lyophilizeHPC-MA in deionized water to form a 15% (w/v) aqueous solution.The transmittance of the sample was monitored as a function oftemperature at a fixed wavelength of 500 nm, using a UV−visspectrophotometer (Agilent Technologies Cary 60 UVVIS, Australia)with the sample cell temperature controlled with a circulating waterbath. The onset temperature was defined as the temperature at whichsigns of opaqueness were first observed. The transmittance wasmeasured four times and the average value of the measurement wastaken.

Preparation of HPC-MA Hydrogels. HPC-MA was dissolved indeionized water at 15% or 20% (w/v), to which a photoinitiator 2-hydroxy-1-[4-(2-hydroxyethoxy)phenyl]-2-methyl-1-propanone wasadded to a concentration of 4% (w/w) of the polymer weight,equivalent to a final concentration of 0.6% for HPC-MA-15% and 0.8%(w/v) for HPC-MA-20%. The solution was first warmed to 40−45 °Cto allow the photonitiator to dissolve, the solution was then allowed tocool down to room temperature before cross-linking. The solution wascross-linked with UV light (320−405 nm, 200 mW cm−2) at a distanceof approximately 200 mm using an UV cross-linker (Honle UVTechnology, UV-F 400, Germany) for approximately 6 min. Thecross-linked gels were then washed with deionized water to removeany uncross-linked HPC-MA and the photoinitiator to minimizepossible cytotoxicity effect from residual photoinitiator. The hydrogelprepared from 15% and 20% (w/v) of HPC-MA are denoted as HPC-MA-15% and HPC-MA-20%, respectively.

Water Contact Angle Measurement. Static contact angles weremeasured using the static sessile drop method by employing watercontact angle equipment (OCA 20, DataPhysics Instrument, GmbH,Germany) equipped with automatic dispenser. In this experiment, a 10μL deionized water droplet was dropped on the surface of the gel andthe angle is measured by using the circle fit. To examine thedependence of contact angle on temperature, 25, 37, and 42 °C wereselected as the testing temperatures.

Equilibrium Swelling Measurement. The as-prepared hydrogelswere submerged in water at 4, 25, 37, and 42 °C for 48 h. The weightof the swollen sample (Wh) was measured after wiping off the excesswater with a piece of filter paper. The weight of the dried sample (Wd)was measured after drying the swollen sample in an vacuum oven at 40°C for 48 h. The weight of samples dehydrated using vacuum oven at40 °C for 48 h were compared to the weight of samples prepared withthe same formulation but dehydrated by lyophilization for 5 days; nosignificant weight difference was noted. The swelling ratio, SR, of thehydrogel was calculated according to eq 1. Each sample was measuredthree times from the four parallel specimens, and the average value ofthe measurement was taken.

=−W W

WSR h d

d (1)

Dynamic Mechanical Analysis. The storage moduli of HPC-MA-15% and HPC-MA-20% were tested on a dynamic mechanical analyzer(DMA) add in software (TA Instruments, Q800) at a frequency of 1Hz and a preload of 0.01 N over a temperature range of 25−45 °Cunder compression mode. The testing was performed underatmospheric condition with a heating rate of 1 °C min−1.

Cell Culture. All culture mediums used were obtained from Gibco,Australia, unless stated otherwise. The Oct4b2 cell lines that possess a

Table 1. Summary of Different Substrates Used for CellSheet Engineering

substrate material release mechanism ref

poly(N-isopropylacrylamide) (PIPAAm) thermoresponsive 48hydroxybutyl chitosan thermoresponsive 16poly(N-isopropylacrylamide)/claynanocomposite hydrogel

thermoresponsive 17

methylcellulose/PBS/collagen hydrogel thermoresponsive 18elastic protein based polymer thermoresponsive 20poly(N-isopropylacrylamide-co-acrylic acid)-b-poly(L-lactic acid)

thermoresponsive 19

RGD-modified poly(L-lysine)-graf t-poly(ethylene glycol)

electrochemicalpolarization

21

poly(N-isopropylacrylamide)-nitrospiropyran photoresponsive 22salmon atelocollagen fibrillar gel collagenase digestion 23

ACS Applied Materials & Interfaces Research Article

dx.doi.org/10.1021/am4009133 | ACS Appl. Mater. Interfaces 2013, 5, 5592−56005593

green fluorescent protein (GFP) under the control of the Oct4promoter (Oct4-GFP)29 were cultured according to the procedurereported in Sarvi et al.30 Prior to cell seeding, Oct4b2 cells werecultured on 0.1% gelatin-coated dish with knockout mediumsupplemented with 20% knockout serum replacement, 1% nones-sential amino acids, 0.1 mM 2-mercaptoethanol, 2 mM Glutamax, 1%penicillin-streptomycin (P/S), and 1000 U/mL murine leukemiainhibitory factor (mLIF, Chemicon, Australia). COS-7 cells weregrown and maintained in Dulbecco’s modified Eagle’s medium(DMEM) supplemented with 10% fetal bovine serum (FBS), 2 mML-glutamine, and 50 units/mL penicillin-streptomycin (P/S). BothCOS-7 and Oct4b2 cells were cultured in a 37 °C, humidified 5% CO2incubator.Temperature-Dependent Cell Release. HPC-MA-20% scaffolds

with dimensions of 1.6 cm diameter by 0.5 cm thickness were sterilizedby soaking in 70% ethanol overnight followed by washing with sterilephosphate buffered saline (PBS) before cell seeding. COS-7 andOct4b2 cells were seeded with a cell density of 2 × 104 cells perscaffold. 10% (v/v) of gelatin was added to the culture media toimprove cell adhesion. The plate was incubated at 37 °C in ahumidified 5% CO2 incubator for 3 h for cell attachment, and then 1mL of fresh culture medium was added and the cells were allowed tocultivate on the hydrogel. COS-7 and Oct4b2 cells detachment wascarried out at selected time point by incubating the culture plates at 4°C for 30 and 10 min, respectively. Culture medium containing thedetached cells was transferred to a new well-plate for reculturing todetermine the regrowth ability of the retrieved cells. Images of cells ateach stage of the process were visualized using fluorescent microscopy(Nikon, Eclipse Ti).Cell Proliferation Assay. The viability of the recultured COS-7

cells was analyzed using the alamarBlue assay on days 1 to day 4. Ateach time point, 200 μL of 10% alamarBlue dye (diluted withcomplete medium) was added to each well containing cold treatedcells and each well containing control cells (cells cultured on well-platewithout cold treatment). The well-plate was then incubated at 37 °Cfor 4 h in a humidified 5% CO2 incubator. Fluorescence analysis usingan excitation wavelength of 570 nm and emission wavelength of 600nm was carried out with a microplate reader (SynergyTM Mx, Biotek).The viability of cells was expressed as the percentage of alamarBluereduction. Cells that have not been cultured on hydrogel and have notsubjected to cold treatment were cultured on a well-plate and used ascontrol.The number of cells proliferating on the hydrogel was determined

by quantifying the DNA content using PicoGreen assay kit (Quanti-iTPicoGreen dsDNA Reagent, Life Technologies, Australia) according tothe manufacturer protocol. In brief, the cells were trypsinized fromsubstrate and centrifuge for 5 min at 1200 rpm. The cell pellet wasrinsed by cold PBS twice each followed by centrifugation for 5 min at

1200 rpm. The resultant cell pellet was collected by discarding thesupernatant. The cells were lysed using 100 μL of NP40 cell lysisbuffer (Life Technologies, Australia) for 30 min on ice and vortexed at10 min intervals. The extract was transferred to microcentrifuge tubesand centrifuge at 13 000 rpm for 10 min at 4 °C. The clear lysate wasaliquoted to clean microcentrifuge tube where 100 μL of PicoGreenreagent was added and allowed to incubate for 5 min at roomtemperature in the dark. The final solution was transferred into a 96-well plate, and the fluorescence was read using a microplate reader atexcitation wavelength 480 nm and emission wavelength 520 nm. Thenumber of cells in the sample was determined by correlating the DNAcontent with a DNA standard curve. The DNA standard curve wasdetermined using cell lysates with known number of cells. In a separateexperiment, the COS-7 cells were cultured on hydrogel followed bycold treatment on day 4, the harvested cells were readhered on culturedish and allowed to proliferate for 5 days. To examine the proliferationof the cold treated cells, COS-7 cells were first cultured on thehydrogels for 4 days followed by 4 °C cold treatment, the harvestedcells were transferred to culture dish without further trypsinization,these cells were allowed to culture for another 5 days, and the DNAcontents were quantified using the PicoGreen assay as describedabove.

Statistical Analysis. All experiments were performed in at leastthree replicates. Results were reported as average value ± standarddeviation. One-way analysis of variance (ANOVA) was used tocompare multiple groups of data statically; p values lower than 0.05were considered statistically significant.

■ RESULTS AND DISCUSSIONPreparation of HPC-MA Macromonomer. HPC-MA was

synthesized using a scheme illustrated in Scheme 1. DCC hasbeen used for activation of carboxylic acids for coupling ofcellulose derivatives with DMAP as catalyst.31,32 DCC is knownas a condensation reagent for producing anhydrides fromcorresponding acids with the formation of urea byproduct. Thelimitation of using anhydride is that only half the acid isstoichiometrically coupled and the other half is wasted33 (i.e.,released in the form of less reactive acid form). The use ofDCC here is to react with these acids to regenerate anhydridestherefore improving the reaction efficacy. The degree ofsubstitution (DS) was determined to be ∼0.4 by 1H NMRspectroscopy. HPC-MA with 0.4 DS was selected as an examplefor scaffold preparation as the higher degree of substitutionpolymer tend to lose its solubility in water and lower degree ofsubstitution fails to provide adequate mechanical properties.Aqueous HPC exhibits low critical solution temperature

Scheme 1. Synthesis Scheme of HPC-MA

ACS Applied Materials & Interfaces Research Article

dx.doi.org/10.1021/am4009133 | ACS Appl. Mater. Interfaces 2013, 5, 5592−56005594

(LCST) transition, from isotropic solutions at room temper-ature to metastable colloidal systems upon heating.24,26 Thisthermal-induced phase separation is due to dehydration ofHPC and consequently increased hydrophobic associationsamong HPC molecules with increasing temperature. HPC-MAretains the phase behavior characteristic of HPC, asdemonstrated in the turbidity measurement shown in Figure1 (p < 0.05, n = 4). The first sign of turbidity was observed at

approximately 36 °C. The LCST of HPC-MA was determinedto be approximately 37 to 38 °C, after which the transmittancedecreases drastically with increasing temperature until reachinga plateau at ∼45 °C.Physiochemical Characterization of HPC-MA Hydro-

gels. UV-irradiation was selected to induce cross-linking ofaqueous HPC-MA due to its many advantages, including fastpolymerization rate at low temperature and ease of hydrogelshaping.34 The photoinit iator , 2-hydroxy-1-[4-(2-hydroxyethoxy)phenyl]-2-methyl-1-propanone (Irgacure2959), has shown to cause minimal toxicity to a wide rangeof mammalian cells35 and, therefore, was employed for thepreparation of HPC-MA-15% and HPC-MA-20% hydrogelsamples. Figure 2 shows the representative photos of the watercontact angles on HPC-MA-20% at 25, 37, and 42 °C, as thetesting temperature is increased from 25 to 37 °C; the watercontact angle of the HPC-MA gel increased from 35.1° to52.1°, reflecting an increase in surface hydrophobicity. Thisfinding revealed that the thermoresponsive behavior of HPC-MA conjugates was retained after cross-linking, resulting inchemically cross-linked thermoresponsive hydrogels that under-

went phase transition at an onset temperature of 37 °C. Afurther increased in temperature above the HPC-MA gel LCSTresults in a contact angle of 61°.Equilibrium swelling of HPC-MA hydrogels was measured

gravimetrically at 4, 25, 37, and 42 °C. Figure 3 shows that as

the temperature increases, the equilibrium swelling ratiodecreases (p < 0.05, n = 5), indicating that the hydrogelsundergo a temperature-triggered deswelling process. Themagnitude of change in equilibrium swelling ratio of thehydrogels, from a swollen to deswollen state, decreases as thepolymer concentration in the hydrogels increases from 15% to20% (w/v). Concurrently, the hydrogels also change theirappearances from transparent to opaque as the temperaturesrises above 42 °C (Figure 4).

Figure 1. Turbidimetry measurement for HPC-MA-15% (black □)and HPC-MA-20% (gray ◇).The percent transmittance decreases asthe temperature increases (p < 0.05, n = 4). Data are shown as meanvalues with standard deviation as error bars in the form of mean value± standard deviation.

Figure 2. Contact angle of HPC-MA-20% gel at room temperature (A), 37 °C (B), and 42 °C (C). Data are shown as mean values with standarddeviation as error bars in the form of mean value ± standard deviation.

Figure 3. Swelling ratio for HPC-MA-15% and HPC-MA-20% at 4 °C(black bar), 25 °C (dark gray bar), 37 °C (medium gray bar), and 42°C (light gray bar). The swelling ratio decreases as the temperatureincreases (p < 0.05, n = 5). Data are shown as mean values withstandard deviation as error bars in the form of mean value ± standarddeviation.

Figure 4. Thermal sensitivity of HPC-MA gel (scale bar of 1 cm) at 25°C (A) and 42 °C (B).

ACS Applied Materials & Interfaces Research Article

dx.doi.org/10.1021/am4009133 | ACS Appl. Mater. Interfaces 2013, 5, 5592−56005595

HPC-MA contains both hydrophilic and hydrophobicgroups. At temperature below the lower critical solutiontemperature, the hydrophilic groups of the HPC-MA in thehydrogels bond to water molecules through hydrogen bonds toform a stable shell around the hydrophobic groups, hencegiving rise to a high equilibrium swelling ratio and low watercontact angle. As seen from Figure 3, the equilibrium swellingratio starts to decrease at approximately 37 °C. This is due tobreaking of the hydrogen bonds between the water moleculesand/surrounding the polymer at elevated temperature, andconsequently the interactions among the hydrophobic groupsbecome dominant. As a result, the water molecules are releasedout, and the polymer networks in the hydrogels collapse withincreased hydrophobicity, similar behavior has been observedfor other types of thermo-responsive hydrogel systems.36,37

Storage moduli of HPC-MA-15% and HPC-MA-20% areshown in Figure 5 as a function of temperature. It is well-

known that cross-linking density in hydrogel increases withincreasing polymer concentration due to the formation oflonger attached chain,4,38 and a densely cross-linked networkexhibits higher mechanical strength compared to loosely cross-linked network. As expected, the storage modulus of thehydrogels increases with increasing the polymer concentration

from 15% to 20% (w/v). As temperature increases from 25 to45 °C, the storage modulus increases from 0.7 to 0.9 kPa forHPC-MA-15%, and from 1.9 to 2.5 kPa from HPC-MA-20%,respectively. The HPC-MA hydrogels exhibit temperaturedependent storage modulus; a similar trend has also beenreported for other types of thermoresponsive hydrogels, such asa hydrogel composed of PNIPAAm and methylcellulose.39 Asthe temperature rises above the LCST, the polymer chainswithin hydrogel undergo coil−globule transition, leading to amore densely packed network, thereby exhibiting higher storagemodulus. The mechanical property of a hydrogel is animportant factor for promoting tissue repair and regener-ation.40,41 Different targeted tissue types require the use ofhydrogels with matching mechanical properties for cellcultivation. The range of storage modulus exhibited by theHPC-MA hydrogels is comparable to that of soft tissues andorgans such as brain, lymph node, mammary gland, liver, breasttumor, and kidney,42 which suggests that they can be used asbiomimicry culture substrates to simulate these soft tissues.Further fine-tuning of the mechanical properties of HPC-MAhydrogels can be achieved by varying the concentration anddegree of substitution of HPC-MA.

Cell Proliferation on HPC-MA Hydrogels. To assess thecytocompatibility of the HPC-MA hydrogels, COS-7 cells wereseeded onto the surface of HPC-MA-20% hydrogel as shown inFigure 6A and cultured at 37 °C in a humidified 5% CO2incubator. Figure 6B−E shows that COS-7 cells exhibitedspindle morphology and reached confluency after 5 days,indicating that these cells adhered well on the hydrogel. Thenumber of cells proliferated on the hydrogel was quantified bydetermining the DNA content, because DNA is the cellularcomponent that reflects the cell number most accurately.43

Figure 7 shows that the DNA content, and hence the numberof cells, on the hydrogel increased over the 5 day period (p <0.05, n = 4), indicating that the HPC-MA hydrogel isbiocompatible to COS-7 cells and facilitate cell proliferation.

Temperature-Dependent Cell-Release Behavior. Tra-ditionally, to detach cells that adhere to culture surfaces,enzymatic digestion such as trypsin or Dispase is utilized. Such

Figure 5. Storage modulus for HPC-MA-15% (black ◆) and HPC-MA-20% (gray ■).

Figure 6. COS 7 cell growth (scale bar of 100 μm): day 0 (A), day 2 (B), day 3 (C), day 4 (D), day 5 (E) on hydrogel, and the surface of hydrogelafter cold treatment (F).

ACS Applied Materials & Interfaces Research Article

dx.doi.org/10.1021/am4009133 | ACS Appl. Mater. Interfaces 2013, 5, 5592−56005596

detachment of cells caused the destruction of extracellularmatrix, growth factor receptors, and cellular junctions.1,5,6 Inthis study, COS-7 cells are employed for monitoring cellattachment on and detachment from the HPC-MA hydrogelsurfaces, because of their rapid division and growth.44 Cellswere seeded onto the surface of HPC-MA-20% hydrogel. Atday 4, the hydrogel was subjected to cold treatment at 4 °C for30 min. It was found that approximately 80 to 90% of theattached cells were detached from the hydrogel by this coldtreatment (Figure 6F). At the onset LCST, cells adhere on thehydrogel surface passively due to hydrophobic interactionbetween the cells and the gel surface. During cold treatment,the hydrogel undergoes phase transition and becomes hydro-philic which is less favorable to cell adhesion, thus causing thecells to detach from the hydrogel surface. The detachment ofcells is accelerated probably due to water penetration as thehydrogel underwent low temperature induced swelling. Inanother set of experiment, COS-7 cells were allowed tocultivate for 10 day (Figure 8A) followed by cold treatment.

Figure 8B shows that a detached COS-7 cell sheet wastransferred and readhered on a culture dish after cold treatmentwithout trypsin treatment. It was observed that cell−cellconnections were maintained after cold detachment, which isimportant for tissue regeneration.Similar process involving cell seeding, attachment and

detachment was carried out using Oct4b2 cells. Figure 9A−Dshows the cells seeded on hydrogel at day 0 and cell growth onhydrogel at day 3. At day 3, the cultured Oct4b2 cells weredetached from the hydrogel followed by incubation at 4 °C for10 min. Reculturing of the detached Oct4b2 cells onto a new

tissue-culture plate was undertaken to preliminarily access theviability of the detached cells. Oct4b2 cells carry the Oct4-greenfluorescence protein (Oct4-GFP) reporter, and the Oct4-GFPexpression is directly correlated with the pluripotency of thesecells. During the course of the experiment, the GFP expressionof Oct4b2 cells was monitored using a fluorescence microscope.Figure 9E and F shows that the detached Oct4b2 cellsdemonstrated good viability and healthy morphology of theOct4b2 cells as compared to Oct4b2 cells cultured on T-flaskwithout cold treatment (control) in Figure 10. The Oct4-GFPexpression observed in Oct4b2 cells at day 3 in Figure 9G andH revealed that the pluripotency of Oct4b2 cells weremaintained after the cold treatment of the hydrogel, indicatingthat these cells can be further proliferated for tissue engineeringapplications.

Cell Viability after Cold Treatment. COS-7 cells werecultured on the hydrogel as described above. At day 4, thehydrogel was subjected to cold treatment at 4 °C for 30 min.

Figure 7. Proliferation of COS-7 cells on HPC-MA-20% hydrogel.Cell growth on hydrogel increased gradually over a period of 5 days (p< 0.05, n = 4). Data are shown as mean values with standard deviationas error bars in the form of mean value ± standard deviation.

Figure 8. (A) COS 7 cells grown on hydrogel (scale bar 0.2 cm) at day10. The hydrogel was subjected to cold treatment; the cell sheet wasthen transferred to a culture dish. (B) Representative image showing acell sheet retrieved after cold treatment (scale bar 100 μm).

Figure 9. Representative images showing bright field and GFPexpression of Oct4b2 cells growth (scale bar of 100 μm) at day 0 (A,B), and day 3 on hydrogel (C, D). Oct4b2 cells were cultured on ahydrogel for a period of 3 days, the hydrogel was then subjected tocold treatment, and the detached cells were transfer to a culture dishfor readhesion. Representative images show the (E) bright field and(F) GFP filter view of the surface of hydrogel after cold treatment;apart from debris, very little cells and GFP signal can be detected,indicating that the cold treatment can detach most of the cells fromthe hydrogel. Representative images show the (G) bright field and (H)GFP filter view of cells retrieved after cold treatment; images weretaken on day 3 after the retrieved cells were transferred to a culturedish.

ACS Applied Materials & Interfaces Research Article

dx.doi.org/10.1021/am4009133 | ACS Appl. Mater. Interfaces 2013, 5, 5592−56005597

The readhesion and proliferation of the detached COS-7 cellswere assessed preliminarily by reculturing them onto a well-plate and measured using alamarBlue assay. Figure 11A shows a

photo of the detached cell sheet from the hydrogel after coldtreatment. The harvested cell sheet shrunk due to the inherentcontractile force within cells and their connectivity to the sheet,similar phenomena was observed by other study that retrievecell sheet from poly(NIPAAm) substrate.45 Hirose et al.46

demonstrated that the use of a hydrophilically modifiedpolyvinylidenefluoride membrane as a supporter can preventthe cell sheet from shrinkage and wrinkling when transferringthe cell sheet to another surface, such support membrane canbe used in future study to prevent the shrinkage of theharvested cell sheet. Nevertheless, once the cell sheet from

HPC-MA hydrogel was allowed to readhere on a culture dish,the cell sheet started to spread out as the cells started to adhereand migrate along the culture dish (Figure 11B−D). It wasshown that these cells adhered well to the new surface andresumed normal growth (Figures 11 and 12). Figure 12 shows

the viability of cells after undergoing cold treatment ascompared to cells that have not been cultured on hydrogeland have not been subjected to cold treatment (control). Themetabolic activity of cells causes a chemical reduction of thealamarBlue reagent and leads to color change in this redoxindicator. In another word, the increase in metabolic activityincreases the amount of reduced alamarBlue.47 The resultshows that there is no significant difference in percentage ofalamarBlue reduction in both the cold treated cells and controlcells (p < 0.05, n = 3). The increase in percentage of alamarBluereduction from ∼3.5% to ∼46% indicates that the cells are not

Figure 10. Bright field and GFP expression of Oct4b2 cells growth on culture dish (scale bar of 100 μm): day 1 (A, B), day 2 (C, D), and day 3 (E,F).

Figure 11. COS 7 cells were cultured on hydrogel for 4 days followedby cold treatment. The detached cells were allowed to readhere andproliferate on culture dish for (a) 0 day, (B) 1 day, (C) 3 days, and(D) 4 days (scale bar of 100 μm).

Figure 12. Cell viability of COS-7 cells with cold treatment (gray bar)compared to cells without hydrogel and cold treatment (black bar)(control) in terms of percent reduction in alamarBlue. No significantdifference in viability of cold treated cells and control cells wasobserved (p < 0.05, n = 3). Data are shown as mean values withstandard deviation as error bars in the form of mean value ± standarddeviation.

ACS Applied Materials & Interfaces Research Article

dx.doi.org/10.1021/am4009133 | ACS Appl. Mater. Interfaces 2013, 5, 5592−56005598

only viable but also proliferate continuously, thus revealing thatthe cells were not affected by undergoing cold treatment ascompared to control cells. This indicates that the cells retrievedfrom the hydrogel can be further proliferated for tissueengineering applications. After cold treatment, the number ofcells was quantified by measuring the DNA content. Figure 13

shows that the cell number increases over a period of 5 days (p< 0.05, n = 6). The growth of cold-treated cells was observed tobe slower compared to those grew on hydrogel before coldtreatment. This is probably due to uneven cell seeding, as thecold-treated cells were readhered on a culture plate as cell sheetwithout further trypsinization, the proliferation of cells at thecenter of a cell sheet is restricted due to the lack of space. Asshown in Figure 11, those cells at the center of the cell sheetneed to migrate out first before proceeding to proliferation.Nonetheless, the increase in cell number confirmed that cellsretained their proliferation ability after cold treatment.The findings from this study suggest that the HPC-MA

hydrogel possess thermoresponsive properties and exhibittemperature dependent cell-release behavior. It is worthwhileto mention that the LCST of HPC-MA can be furthermodulated by engineering the side chain chemistry. Theinfluence of introduced side groups on the thermal-responsiveproperties of HPC-MA hydrogel is currently under inves-tigation.

■ CONCLUSIONServing as a proof of concept study, thermoresponsive cellulosichydrogels were prepared by UV cross-linking aqueous HPC-MA at room temperature. The hydrogels exhibit temperature-modulated characteristics including surface hydrophilicity andhydrophobicity, equilibrium swelling and deswelling, as well asmechanical properties. A LCST of HPC-MA conjugate wasfound to be ∼37−38 °C as determined by turbiditymeasurement. Cell-releasing characteristics were demonstratedusing COS-7 cells and Oct4b2 cells. Both COS-7 and Oct4b2cells adhere and proliferate well on the hydrogel surfaces undernormal cell-culture conditions. By reducing the temperature to4 °C, the cultivated cells spontaneously detach from thehydrogels without trypsin treatment. These unique properties

of HPC-MA hydrogels would make them potential culturesubstrates for cell sheet engineering.

■ AUTHOR INFORMATION

Corresponding Author* (P.P.Y.C.) Tel: +613 9925 2660. E-mail: peggy.chan@rmit.edu.au. (Z.Y.) Tel: +612 4221 3832. E-mail: zyue@uow.edu.au.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

Funding for this research was partly provided through anAustralia Research Council Discovery Project Grant ARC DP120102570. We also thank Dr. Paul Verma from MonashInstitute of Medical Research (MIMR) for supplying Oct4b2cells. We would like to thank Prof. Robert Shanks and IzanRoshawaty Mustapa for their assistance with DMA measure-ments. This work was performed in part at the MelbourneCentre for Nanofabrication, an initiative partly funded by theCommonwealth of Australia and the Victoria Government.P.P.Y.C. is grateful for a MCN Technology Fellowship and aRMIT University Senior Research Fellowship that supportedthis work.

■ REFERENCES(1) Shimizu, T.; Yamato, M.; Kikuchi, A.; Okano, T. Biomaterials2003, 24 (13), 2309−2316.(2) Iwata, T.; Yamato, M.; Okano, T. Intelligent Surfaces for Cell-Sheet Engineering. In Principles of Regenerative Medicine, second ed.;Atala, A., Lanza, R., Thomson, J., Nerem, R., Eds.; Academic Press: SanDiego, 2011; Chapter 29, pp 517−527.(3) Zahn, R.; Thomasson, E.; Guillaume-Gentil, O.; Voros, J.;Zambelli, T. Biomater 2012, 33 (12), 3421−7.(4) Lee, F.; Chung, J. E.; Kurisawa, M. J. Controlled Release 2009, 134(3), 186−193.(5) Biazar, E.; Montazeri, N.; Pourshamsian, K.; Asadifard, F.;Ghorbanalinezhad, E.; Keshel, S. H.; Hashemi, M.; F.H, S. R.; Majdi,A. J. Paramed. Sci. 2010, 1 (3), 27−33.(6) Yamato, M.; Akiyama, Y.; Kobayashi, J.; Yang, J.; Kikuchi, A.;Okano, T. Prog. Polym. Sci. 2007, 32 (8−9), 1123−1133.(7) da Silva, R. M. P.; Mano, J. F.; Reis, R. L. Trends Biotechnol. 2007,25 (12), 577−583.(8) Wu, J.-Y.; Liu, S.-Q.; Heng, P. W.-S.; Yang, Y.-Y. J. ControlledRelease 2005, 102 (2), 361−372.(9) Okano, T. Rinsho Shinkeigaku 2006, 46 (11), 795−798.(10) Biazar, E.; Pourshamsian, K. Orient. J. Chem. 2011, 27 (4),1443−1449.(11) Haraguchi, Y.; Shimizu, T.; Yamato, M.; Okano, T. Cardiol. Res.Practice 2011, 2011, 1−8.(12) Masuda, S.; Shimizu, T.; Yamato, M.; Okano, T. Adv. DrugDelivery Rev. 2008, 60 (2), 277−285.(13) Sakaguchi, K.; Shimizu, T.; Horaguchi, S.; Sekine, H.; Yamato,M.; Umezu, M.; Okano, T. Sci. Rep. 2013, 3, 1316.(14) Wang, T.; Liu, D.; Lian, C.; Zheng, S.; Liu, X.; Wang, C.; Tong,Z. React. Funct. Polym. 2011, 71 (4), 447−454.(15) Vihola, H.; Laukkanen, A.; Valtola, L.; Tenhu, H.; Hirvonen, J.Biomater 2005, 26 (16), 3055−3064.(16) Chen, B.; Dang, J.; Tan, T. L.; Fang, N.; Chen, W. N.; Leong, K.W.; Chan, V. Biomater 2007, 28 (8), 1503−1514.(17) Haraguchi, K.; Takehisa, T.; Ebato, M. Biomacromolecules 2006,7 (11), 3267−3275.(18) Chen, C.-H.; Tsai, C.-C.; Chen, W.; Mi, F.-L.; Liang, H.-F.;Chen, S.-C.; Sung, H.-W. Biomacromolecules 2006, 7 (3), 736−743.(19) Kim, Y. S.; Lim, J. Y.; Donahue, H. J.; Lowe, T. L. Tissue Eng2005, 11 (1−2), 30−40.

Figure 13. Proliferation of COS-7 cells that were harvested fromhydrogel using cold treatment. COS 7 cells were firsts cultured onhydrogel for 4 days followed by cold treatment; the detached cellswere allowed to readhere and proliferate on culture dish. The numberof cells increased gradually over a period of 5 days (p < 0.05, n = 6).The data are shown as mean values with standard deviation as errorbars in the form of mean value ± standard deviation.

ACS Applied Materials & Interfaces Research Article

dx.doi.org/10.1021/am4009133 | ACS Appl. Mater. Interfaces 2013, 5, 5592−56005599

(20) Zhang, H.; Iwama, M.; Akaike, T.; Urry, D. W.; Pattanaik, A.;Parker, T. M.; Konishi, I.; Nikaido, T. Tissue Eng. 2006, 12 (2), 391−401.(21) Guillaume-Gentil, O.; Akiyama, Y.; Schuler, M.; Tang, C.;Textor, M.; Yamato, M.; Okano, T.; Voros, J. Adv. Mater. 2008, 20 (3),560−565.(22) Edahiro, J.-i.; Sumaru, K.; Tada, Y.; Ohi, K.; Takagi, T.; Kameda,M.; Shinbo, T.; Kanamori, T.; Yoshimi, Y. Biomacromolecules 2005, 6(2), 970−974.(23) Nagai, N.; Yunoki, S.; Satoh, Y.; Tajima, K.; Munekata, M. J.Biosci. Bioeng. 2004, 98 (6), 493−496.(24) Yue, Z.; Wen, F.; Gao, S.; Ang, M. Y.; Pallathadka, P. K.; Liu, L.;Yu, H. Biomaterials 2010, 31 (32), 8141−521.(25) Hussain, M. A. J. Polym. Sci., Part A: Polym. Chem. 2008, 46 (2),747−752.(26) Hirsch, S. G.; Spontak, R. J. Polymer 2002, 43 (1), 123−129.(27) Francis, M. F.; Piredda, M.; Winnik, F. M. J. Controlled Release2003, 93 (1), 59−68.(28) Hoo, S. P.; Loh, Q. L.; Yue, Z.; Fu, J.; Tan, T. T. Y.; Choong, C.;Chan, P. P. Y. J. Mater. Chem. B 2013, 1 (24), 3107−3117.(29) Yeom, Y. I.; Fuhrmann, G.; Ovitt, C. E.; Brehm, A.; Ohbo, K.;Gross, M.; Hubner, K.; Scholer, H. R. Development 1996, 122 (3),881−894.(30) Sarvi, F.; Yue, Z.; Hourigan, K.; Thompson, M. C.; Chan, P. P.Y. J. Mater. Chem. B 2013, 1 (7), 987−996.(31) Yue, Z.; Cowie, J. M. G. Polymer 2002, 43 (16), 4453−4460.(32) Zhang, Z. B.; McCormick, C. L. J. Appl. Polym. Sci. 1997, 66 (2),293−305.(33) Montalbetti, C. A. G. N.; Falque, V. Tetrahedron 2005, 61 (46),10827−10852.(34) Benson, R. S. Nucl. Instrum. Methods Phys. Res., Sect. B 2002, 191(1−4), 752−757.(35) Nair, L. S.; Laurencin, C. T.; Tandon, M. Injectable Hydrogelsas Biomaterials. In Advanced Biomaterials; Basu, B., Kati, D. S., Kumar,K., Eds.; John Wiley & Sons, Inc.: Hoboken, NJ, 2010; pp 179−203.(36) Cole, M. A.; Voelcker, N. H.; Thissen, H.; Griesser, H. J.Biomaterials 2009, 30 (9), 1827−1850.(37) Temtem, M.; Casimiro, T.; Mano, J. F.; Aguiar-Ricardo, A.Green Chem. 2007, 9 (1), 75−79.(38) Wang, L. S.; Chow, P. Y.; Phan, T. T.; Lim, I. J.; Yang, Y. Y. Adv.Funct. Mater. 2006, 16 (9), 1171−1178.(39) Liu, W.; Zhang, B.; Lu, W. W.; Li, X.; Zhu, D.; De Yao, K.;Wang, Q.; Zhao, C.; Wang, C. Biomaterials 2004, 25 (15), 3005−3012.(40) Nicodemus, G. D.; Bryant, S. J. Tissue Eng., Part B 2008, 14 (2),149−65.(41) Straley, K. S.; Foo, C. W.; Heilshorn, S. C. J. Neurotrauma 2010,27 (1), 1−19.(42) Levental, I.; Georges, P. C.; Janmey, P. A. Soft Matter 2007, 3(3), 299−306.(43) Rage, R.; Mitchen, J.; Wilding, G. Anal. Biochem. 1990, 191 (1),31−34.(44) Knapek, K.; Frondorf, K.; Post, J.; Short, S.; Cox, D.; Gomez-Cambronero, J. Mol. Cell. Biol. 2010, 30 (18), 4492−506.(45) Mizutani, A.; Kikuchi, A.; Yamato, M.; Kanazawa, H.; Okano, T.Biomaterials 2008, 29 (13), 2073−2081.(46) Hirose, M.; Kwon, O. H.; Yamato, M.; Kikuchi, A.; Okano, T.Biomacromolecules 2000, 1 (3), 377−381.(47) Das, G. K.; Chan, P. P. Y.; Teo, A.; Loo, J. S. C.; Anderson, J.M.; Tan, T. T. Y. J. Biomed. Mater. Res., Part A 2010, 93A (1), 337−346.(48) Kushida, A.; Yamato, M.; Konno, C.; Kikuchi, A.; Sakurai, Y.;Okano, T. J. Biomed. Mater. Res. 2000, 51 (2), 216−223.

ACS Applied Materials & Interfaces Research Article

dx.doi.org/10.1021/am4009133 | ACS Appl. Mater. Interfaces 2013, 5, 5592−56005600